A new venture between the Mayo Clinic and a Boston-area artificial intelligence firm will work with drug companies to discover whether molecules researched for treatment of one disease can effectively treat other conditions.

The month-old research company is called Qrativ and will target conditions for which current treatments are lacking. Mayo's partner in the joint venture is Cambridge-based Nference.

"Mayo's interest in this is not only will we be making new therapies for patients, but I believe that this is also going to enhance Mayo's research capabilities," said Mayo's Dr. Andrew Badley, an infectious disease specialist who directs Mayo's Office of Translation to Practice. "Already we've used the Nference platform for a few investigators at Mayo, which has led to new insights and new associations that have advanced the progress of science tangibly."

Nference is bringing the advanced computing power of its Darwin.ai artificial neural network to bear on huge sets of drug and disease data.

Mayo experts will evaluate the results to see whether the novel therapies suggested by the computer could be particularly useful in treating rare diseases, cancers and other conditions.

Qrativ (pronounced CURE-ative) has completed an $8.3 million round of fundraising and anticipates raising additional money. Mayo Clinic and private equity firms Matrix Capital Management and Matrix Partners have invested in the effort.

Mayo will use any profits from the venture to support its nonprofit goals in health care, medical education and research.

Repurposing old drug compounds for new uses is not new. One of the best-known examples is thalidomide, which was banned in the 1960s as a treatment for morning sickness because it caused birth defects and fetal deaths but was eventually approved in the United States in 2006 as a safe and effective treatment for the blood cancer multiple myeloma.

But the repurposing of old drugs typically happens in a haphazard way. Qrativ wants to use artificial intelligence to systematically evaluate large numbers of drug molecules against a long list of medical conditions. The results of those digital analyses would then be forwarded to medical experts at Mayo to see which conditions have unmet medical needs.

"We prefer the word 'purposed' to 'repurposed,' " Badley said. "We are not just going to take existing drugs and find better homes for them, but will take late-stage preclinical candidates and advance them for the best condition that they can treat."

Qrativ Chief Executive Murali Aravamudan, who is also the CEO of Nference, stressed that the artificial intelligence system would be used only to generate hypotheses from large data sets, and any decision to investigate a new therapy in humans would be made by Mayo experts.

Such therapies would be examined in controlled trials, not used in unapproved "off-label" ways.

Joe Carlson • 612-673-4779